pymongo-pubsub — a publish-subscribe pattern implementation for pymongo

Ever wished you could use your MongoDB to drive an event-driven environment?

Now you can.

pymongo-pubsub uses a round-robin capped collection to store its data on one end and a tailable cursor to catch it as it is inserted on the other.

The publisher

A publisher is a piece of code responsible for pumping data into the database:

import pymongo
from pymongo_pubsub import Publisher

connection = pymongo.Connection()
database = connection.pubsub_db

publisher = Publisher(database, 'test_event')
publisher.push({'message': 'hello world', 'answer': 42})

A Publisher will take care of setting up the underlying collection. You can pass it optional max_size and max_length parameters to define storage limits of the collection at the time it is created. The default size is 1000000 bytes and the default length limit is None.

Make sure your queue is big enough to never overflow or you're going to start losing data. You care about your data, right?

The subscriber

A subscriber is a piece of code responsible for processing data as it is pulled from the database:

import pymongo
from pymongo_pubsub import Subscriber

connection = pymongo.Connection()
database = connection.pubsub_db

def test_cb(data):
    print '{0:f}: {1:s}'.format(data['_id'], data['message'])

subscriber = Subscriber(database, 'test_event', callback=test_cb,
                        matching={'answer': 42})
subscriber.listen()

A Subscriber will never create any collections in the database and in case the collections is missing it will cowardly raise a KeyError exception. This is not a bug. This is to avoid having to synchronize collection limits between the publisher and all of the subscribers.